Abstract Absent new treatment options, bacteria evolve to develop resistance to existing drugs and spread rapidly across the globe. Thus, there is a strong need to identify drugs that: 1) act on bacterial pathogens via novel molecular targets, and 2) are chemically unrelated to clinically used antibiotics. Our innovative approach is designed to meet both of those criteria. Here we propose a novel, general platform to drug discovery that differs from both traditional phenotype- and target-based strategies, and requires a small fraction of the usual costs of drug discovery. Specifically, our proposed drug discovery platform leverages recent advances in bacterial systems biology and builds on large-scale analysis of the mechanisms of action of approved and experimental drugs developed for various indications. Three recent technological advances, coupled with the extensive relevant experience of the key personnel, underlie the proposal: (i) next-generation sequencing, transcriptomic, and metabolomic analyses support precise delineation of the mechanism of action for anti-bacterial drugs; (ii) recent commercial implementation of small molecule shape-comparison methods on GPUs dramatically increase the scale of problems amenable to accurate computational chemistry analysis; and (iii) careful curation of public resources and availability of proprietary databases of approved and experimental drugs enable analysis of large sets of clinical molecules with pre-defined properties. The innovative drug discovery platform proposed herein is designed to identify novel targets, and to discover and develop drugs to treat drug-resistant bacterial infections. In order to establish the feasibility of the approach, we propose to develop drugs working via novel mechanisms of action (MOA) against innately antibiotic-resistant Mycobacterium abscessus.